Calcium phosphates are the principal constituent of hard tissues (bone, cartilage, tooth enamel and dentine). Naturally-occurring bone mineral is made of nanometer-sized, poorly-crystalline calcium phosphate with an apatitic structure. The poorly crystalline apatitic calcium phosphate of bone is distinguished from the more crystalline hydroxyapatites and non-stoichiometric hydroxyapatites by its distinctive x-ray diffraction pattern as shown in FIG. 1. Unlike the ideal stoichiometric crystalline hydroxyapatite, Ca.sub.10 (PO.sub.4).sub.6 (OH).sub.2, with atomic Ca/P ratio of 1.67, the composition of bone mineral is significantly different and may be represented by the following formulae, EQU Ca.sub.8.3 (PO.sub.4).sub.4.3 (HPO.sub.4, CO.sub.3).sub.1.7 (OH, CO.sub.3).sub.0.3.
Bone mineral non-stoichiometry is primarily due to the presence of divalent ions, such as CO.sub.3.sup.2- and HPO.sub.4.sup.2-, which are substituted for the trivalent PO.sub.4.sup.3- ions. Substitution by HPO.sub.4.sup.2- and CO.sub.3.sup.2- ions produces a change of the Ca/P ratio, resulting in Ca/P ratio which may vary between 1.50 to 1.70, depending on the age and bony site. Generally, the Ca/P ratio increases during aging of bone, suggesting that the amount of carbonate species typically increases for older bones. It is the Ca/P ratio in conjunction with nanocrystalline size and the poorly-crystalline nature that yields specific solubility property of the bone minerals. And because bone tissues undergo constant tissue repair regulated by the mineral-resorbing cells (osteoclasts) and mineral-producing cells (osteoblasts), solubility behavior of minerals is important in maintaining a delicate metabolic balance between these cell activities.
Synthetic bone graft material made to closely resemble natural bone minerals can be a useful replacement for natural bone. Acceptable synthetic bone can avoid the problem of availability and harvesting of autologous bone (patient's own bone) and the risks and complications associated with allograft bone (bone from a cadaver), such as risks of viral transmission. An ideal synthetic bone graft should possess a minimum of the following four properties: (1) it should be chemically biocompatible; (2) it should provide some degree of structural integrity in order to keep the graft in place and intact until the patient's own bone heals around it; (3) it should be resorbable so that the patient's own bone ultimately replaces the graft; and, (4) because it may be necessary to incorporate cells and/or biomolecules into the synthetic bone material, it is desirable that the process used to form the material employ low temperatures and chemically mild conditions. Similar criteria are also important for other hard tissue grafts (e.g. cartilage).
These criteria may be met by a material in which parameters, such as Ca/P ratios, crystal size, crystallinity, porosity, density, thermal stability and material purity are controlled. While there have been considerable efforts to synthesize a ceramic material for the use as implants, synthetic hydroxyapatites have most often been used because their chemical formulae are very similar to the naturally occurring mineral in bone. LeGeros R. Z., in Calcium Phosphates in Oral Biology and Medicine, Karger Pub. Co., New York, 1991 teaches highly crystalline forms of hydroxyapatite produced by solution precipitation followed by sintering at high temperatures (800-1200.degree. C.). High temperature treatment yields highly stoichiometric hydroxyapatite with crystal sizes on the order of several microns with a Ca/P of 1.67. Such highly crystalline hydroxyapatite is essentially non-resorbable in vivo. It is not replaced by living bone tissue and remains intact in the patient for an undesirably extended period of time.
A number of other approaches to the production of bone substitute material have employed hydroxyapatite produced by a solid-state acid-base reaction of primarily crystalline calcium phosphate reactants. These hydroxy apatite bone substitute materials are sometimes poorly-reacted, inhomogeneous, and have significant crystalline hydroxyapatite content.
Constantz in U.S. Pat. No. 4,880,610 reports on the preparation of calcium phosphate minerals by the reaction of a highly concentrated phosphoric acid with a calcium source in the presence of a base and hydroxyapatite crystals. The resultant product is a polycrystalline material containing a crystalline form of hydroxyapatite minerals. Likewise, U.S. Pat. No. 5,053,212 to Constantz et al. discloses the use of a powdered acid source to improve the workability and mixability of the acid/base mixture; however, a mixed-phase calcium phosphate material similar to that of U.S. Pat. No. 4,880,610 is reported. Recently, Constantz et al. reported in Science (Vol. 267, pp. 1796-9 (Mar. 24, 1995)) the formation of a carbonated apatite from the reaction of monocalcium phosphate monohydrate, alpha-tricalcium phosphate, and calcium carbonate in a sodium phosphate solution, to provide a calcium phosphate material which is still substantially more crystalline in character than naturally occurring bone minerals.
Similarly, Brown et al. in U.S. Reissue No. 33,221 report on the reaction of crystalline tetracalcium phosphate (Ca/P of 2.0) with acidic calcium phosphates. Liu et al. in U.S. Pat. No. 5,149,368 disclose the reaction of crystalline calcium phosphate salts with an acidic citrate.
A number of calcium phosphate bone fillers and cements have been described as "resorbable." Generally, these are compounds comprising or derived from tricalcium phosphate, tetracalcium phosphate or hydroxyapatite. At best these materials may be considered only weakly resorbable. Of these, the tricalcium phosphate compounds have been demonstrated to be the most resorbable and after many years of study they are still not widely used in clinical settings. The tricalcium phosphates are known to have lengthy and somewhat unpredictable resorption profiles, generally requiring in excess of one year for resorption. Furthermore, unless steps are taken to produce extremely porous or channeled samples, the tricalcium phosphates are not replaced by bone. Recently it has been concluded that the "biodegradation of TCP, which is higher than that of Hap [hydroxyapatite] is not sufficient" (Berger et al., Biomaterials, 16:1241 (1995)).
Tetracalcium phosphate and hydroxyapatite derived compounds are also only weakly resorbable. Tetracalcium phosphate fillers generally exhibit partial resorption over long periods of time such as 80% resorption after 30 months (Horioglu et al., Soc. for Biomaterials, Mar. 18-22, pg 198 (1995)). Approximately 30% of microcrystalline HA implanted into the frontal sinus remained after 18 months in cats.
All of these references disclose a chemical reaction resulting in crystalline form of hydroxyapatite solids that has been obtained by reacting crystalline solids of calcium phosphate. There has been little reported on the use of amorphous calcium phosphates (Ca/P of approximately 1.5) as one of the reactants because the amorphous calcium phosphates are the least understood solids among the calcium phosphates and amorphous calcium phosphate has traditionally been considered to be a relatively inert and non-reactive solid.
Amorphous calcium phosphate material has been used as a direct precursor to the formation of a highly crystalline hydroxyapatite compounds under generally high temperature treatments. Such a highly crystalline material is inappropriate for synthetic bone because it is highly insoluble under physiological conditions. Chow et al. in U.S. Pat. No. 5,525,148 report the testing of ACP precursors in a number of reaction schemes but states that slurries of a variety of crystalline calcium phosphates including ACP either alone or in mixtures do not produce a setting cement or act as an effective remineralizing agent.
Brown et al. in U.S. Reissue No. 33,221 report on the formation of crystalline hydroxyapatite for dental cement by reacting an amorphous phase specifically restricted to tetracalcium phosphate (Ca/P of 2.0) with at least one of the more acidic calcium phosphates. Further, Brown et al., does not disclose the preparation or the properties of such a tetracalcium phosphate in amorphous state. Tung in U.S. Pat. No. 5,037,639 discloses the use and application of standard amorphous calcium phosphate paste for the remineralization of teeth. Tung proposes the use of standard inert amorphous calcium phosphate mixed with and delivered through as a chewing gum, mouth rinse or toothpaste, which upon entering oral fluids converts to crystalline fluoride containing hydroxyapatite which is useful to remineralize tooth enamel. Simkiss in PCT/GB93/01519 describes the use of inhibitors, such as Mg ions or pyrophosphate, mixed with amorphous calcium phosphate and implanted into living tissues. Upon leaching of, for example Mg ions, into surrounding bodily fluids, the amorphous calcium-magnesium phosphate converts into a more crystalline form.
There remains a need for a synthetic bone material that more closely mimics the properties of naturally-occurring minerals in bone. In particular, there remains a need to provide synthetic bioceramics which are completely bioresorbable and biocompatible. The use of such a resorbable calcium phosphate in biomedical devices provides many advantages over alternative conventional materials. For instance, it eliminates the need for post-therapy surgery to remove the device and degrades in the human body to biocompatible, bioresorbable products.
Even when the properties of biocompatibility and bioresorbability have been met, the calcium phosphate used in the medical implant may not have the appropriate range of mechanical properties required in a particular application. Biocompatible ceramics have been reported to have excellent compatibility with bone, but the material may lack the necessary strength, flexibility or other mechanical property. Likewise, when used as an adhesive, the material may lack the necessary adhesive strength. Thus, ceramic materials alone, whether superbly biocompatible or not, may not meet all the requirements of an in vivo medical device. Organic polymeric materials have been used as an alternative material because their mechanical properties can be varied as the particular application requires. However, the bone-bonding capability of and/or the bone reossification of organic polymeric materials have not been shown to be as effective as bioactive ceramics.
It is known that composite materials may possess a blend of the attributes and properties of the component materials comprising the composite. Therefore, it is desirable to provide a bioresorbable bone substitute composite material which advantageously possesses superior bioresorbability, biocompatibility and bone reossification capability, while further possessing the high tensile strength, flexibility or other properties otherwise deficient or lacking in the biomaterial alone.
Calcium phosphate composites have been reported, however they lack the complete bioresorbability required of a superior bone substitute material. Calcium phosphates/polymer composites of the prior art generally are prepared with biocompatible polymers. Calcium phosphate fibers or particles are added as a biocompatible, but not bioresorbable, filler to modify the mechanical properties of the polymer matrix. Prior art composites to date have failed to demonstrate a biocompatible composite in which the bioceramic is fully and rapidly resorbable.